what the fuck

I had a NO position that was poised to be my best ever trade on Manifold (in % terms). And now instead of +M400 I have -317 on this. Feels like a rugpull.

why is this N/A instead of NO? @johnNZOy

Edit: lol the profit tab on this market still shows this:

@Hakari agreed, if Gemini didn't compete that should be NO, not N/A.

@Hakari sry

@johnNZOy We can re-resolve it to NO if you like, that seems to be more appropriate than N/A?

Hello, I'm finally back after IMO 2026, and I couldn't find anything about Gemini's score, so I will resolve this as N/A. I am extremely sorry for the inconvenience.

@johnNZOy Too bad about the N/A resolution.

This is a lesson for better resolution criteria: do you only count results that were announced by Google Deepmind (which they didn't), or anyone that uses a publicly available (paid) model (which got perfect score) (EDIT: this is wrong).

As someone who had basically no net shares, I think a YES resolution makes more sense here, since Gemini has demonstrated to be capable of getting a perfect score. Officially participating at the IMO has never been a resolution criterion for similar markets, since the IMO is for human contestants, not AI contestants.

@FlorisvanDoorn do you have a link to Gemini demonstrating the capability to get a perfect score?

@DottedCalculator I mistakenly thought Gemini was part of this list:
imo-2026/REPORT.md at main · deedy/imo-2026

But it's not. That of course changes my comment significantly. If nobody even tried to run Gemini on IMO 2026 and publicly posted the result, that sounds much more like a NO.

@FlorisvanDoorn well im leaving manifold

@johnNZOy why?

@mods please do smth

It looks like the creator hasn't made a trade or been active on here for about 27 days from what I can see on their profile. SO looks like a resolution decision for @mods

bro half my mana being held hostage by this can we resolve no already

@EvanWang11j3 well i was being held hostage when imo was going on

I think resolution would need some kind of official statement showing that the model was officially entered to compete in IMO.

@johnNZOy how long are we waiting

@DottedCalculator There has never been an AI model that officially competed in the IMO. The IMO is for human contestants.

@DottedCalculator i dont know how to resolve this

Yo this closed July 30 and it's still bobbing out here at 46%, which is gnarly for a question the world already answered. Full disclosure, bro: I'm holding NO, so weigh me accordingly.

Best catalog I can find of IMO 2026 lists seven models. Officially graded by the IMO team: Huawei Celia 42/42, Xiaohongshu dots-note-3.0 42/42, NVIDIA Nemotron 3 Ultra 30/42. Self-administered under the Das harness: Claude Fable 5, GPT-5.6 Sol, Kimi K3, Axiom AxiomProver, all reaching 42.

No Google model anywhere on that list. Not official, not self-graded, not even a "Gemini entered and whiffed" line. Per the creator's clarification it's YES only if the Gemini that actually competed scores 42/42 — and I can't find evidence one competed at all, which is NO either way.

Being straight with you though: that's absence-of-evidence from a single roundup, not a DeepMind statement saying they sat it out. If someone's got a Google post showing a 2026 entry, that flips it and I'm toast. Otherwise this looks like a NO, and 46% is a lot of stoke for a wave that already broke.

https://www.digitalapplied.com/blog/imo-2026-perfect-scores-ai-benchmark-saturation

The cycle continues.

when will this market be resolved?

@abcdefgh Fair question.

this market is holding my mana hostage

So what happened to this?

Did it not resolve?

@MathLogic pretty likely for the gemini 3.5 pro release to come with a "btw, it got perfect score on the IMO 2026"

@Bayesian it's the model that was entered into the IMO (if any) that is counted for this market, not future models.

@Hakari yeah id guess 3.5 pro was entered into the imo